Study/Data Info

We are utilizing data from the longitudinal study “Transitions in Adolescent Girls,” (Barendse et al., 2019) specifically focusing on data from baseline (Wave 1). This dataset includes 174 participants who were 10-13 at the baseline. Our variables of interest are emotion regulation strategy use and depressive symptoms. These were measured using the Emotion Regulation Questionnaire (ERQ) (Gross & John, 2003) and the Center for Epidemiological Studies Depression Scale for Children (CES- DC). (Weissman et al., 1980) These are continuous scales which provide total scores for reappraisal and suppression (ERQ) and depressive symptoms (CES-DC).

Topic Overview

Adolescence is a pivotal period of development characterized by rapid physical, social, and emotional change (Guyer et al., 2016; Nelson et al., 2016) Early adolescence has been proposed as a particularly crucial period for the development of emotion regulation (Silvers, 2022), the processes by which individuals moderate and modulate their emotional experiences and expression (Gross, 1998; Gross, 2015). Additionally, adolescents face heightened vulnerability to the depressive symptoms (Kessler et al., 2007), with depression rates spiking between early and middle adolescence and even steeper rates for girls (Hankin et al., 2007).

Present Study

Drawing upon the longitudinal TAG dataset, we sought to examine whether two particularly salient and well-studied emotion regulation strategies (reappraisal and suppression) were associated with depressive symptoms in early adolescent girls (age 10-13).

These findings will serve as a foundation for future studies investigating these relations across time utilizing all six waves of the data.

Research Questions:

Is higher reappraisal use associated with lower depressive symptoms in adolescent girls?
Is higher suppression use associated with higher depressive symptoms in adolescent girls?

Load packages

library(tidyverse)
library(flextable)
library(here)
library(readxl)
library(modelsummary)
library(janitor)
library(performance)

Load data

We first had to load in four separate datasets to join which contained age, race, emotion regulation, and depressive symptoms:

# Demographic
age <- read_csv(here("data", "TAG_age_session_dates_4waves.csv"))
age <- clean_names(age)

race <- read_excel(here("data", "TAG_W1_Race_Ethnicity.xlsx"))
race <- clean_names(race)

# Emotion Regulation(ERQ)
erq <- read_csv(here("data", "ERQ_Wave1.csv"))
erq <- erq %>% 
       clean_names() %>% 
       distinct(tagid, .keep_all = TRUE)

# Adolescent Depression(CESDC)
cesdc <- read_csv(here("data", "CESDC_Wave1.csv"))
cesdc <- clean_names(cesdc)

Participant Demographics

Race/ethnicity category

First we had to clean the race/ethnicity data by identifying subjects reported multiple racial categories and rename them to “Multi-racial”

race_clean <- race %>%
  mutate(
    w1_ethnicity = as.character(w1_ethnicity),
    w1_ethnicity = if_else(
      str_detect(w1_ethnicity, ","),
      "g. Multi-racial",
      w1_ethnicity
    )
  )

Race/ethnicity distribution

race_clean %>%
  count(w1_ethnicity) %>%
  ggplot(aes(x = fct_reorder(w1_ethnicity, n), y = n, fill = w1_ethnicity)) +
  geom_col(show.legend = FALSE) +
  coord_flip() +
  labs(
    title = "Distribution of Race/Ethnicity in Sample",
    x = "Race/Ethnicity",
    y = "Number of Participants"
  ) +
  theme_minimal(base_size = 13)

Participant demographics cont.

Age

Each wave consisted of two sessions, so we had to calculate the average age between sessions:

age <- age %>% 
  mutate(W1_mean_age = (w1s1_age + w1s2_age)/2) 

# Check mean age for sample
age_summary <- age %>%
  summarize(
    mean_W1_age = mean(W1_mean_age, na.rm = TRUE),
    sd_W1_age   = sd(W1_mean_age, na.rm = TRUE),
  )
print(age_summary)

# A tibble: 1 × 2
  mean_W1_age sd_W1_age
        <dbl>     <dbl>
1        11.6     0.808

Joining all datasets

df <- erq %>% 
  left_join(cesdc, by= 'tagid')

df <- df %>% 
     left_join(age, by = 'tagid')

df <- df %>% 
     left_join(race_clean, by = 'tagid') %>% 
     select(tagid, erq_reappraisal_total, erq_suppression_total, ces_dc_total_75perc, W1_mean_age, w1_ethnicity, 
            w1s1_date, w1s1_age, w1s2_date, w1s2_age, w2s1_date, w2s1_age, w2s2_date, w2s2_age, 
            w3s1_date, w3s1_age, w3s2_date, w3s2_age, w4s1_date, w4s1_age, w4s2_date, w4s2_age
            )

Renaming

In order to tidy the data, we first had to rename each wave/session/date/age variable

df <- df %>%
  rename(
    wave1_session1_date = w1s1_date,
    wave1_session1_age  = w1s1_age,
    wave1_session2_date = w1s2_date,
    wave1_session2_age  = w1s2_age,

    wave2_session1_date = w2s1_date,
    wave2_session1_age  = w2s1_age,
    wave2_session2_date = w2s2_date,
    wave2_session2_age  = w2s2_age,

    wave3_session1_date = w3s1_date,
    wave3_session1_age  = w3s1_age,
    wave3_session2_date = w3s2_date,
    wave3_session2_age  = w3s2_age,

    wave4_session1_date = w4s1_date,
    wave4_session1_age  = w4s1_age,
    wave4_session2_date = w4s2_date,
    wave4_session2_age  = w4s2_age
  )

Tidy and filter

# Tidy
df2 <- df %>%
  mutate(across(starts_with("wave"), as.character))

df_long <- df2 %>%
  pivot_longer(
    cols = starts_with("wave"),
    names_to = c("wave", "session", "info"),
    names_sep = "_",
    values_to = "value"
  ) %>%
  mutate(
    wave    = parse_number(wave),
    session = parse_number(session)
  )

df_tidy <- df_long %>%
  pivot_wider(
    names_from = info,
    values_from = value
  ) %>% 
  mutate(age = as.numeric(age))


# Filter
df_final <- df_tidy %>% 
  filter(wave == 1, session == 1)

Results

Cohort Plot

Sample Summary Statistics

df_names <- df_tidy %>% 
    select(-wave, -session) %>% 
    rename(
    Reappraisal       = erq_reappraisal_total,
    Suppression       = erq_suppression_total,
    Depression        = ces_dc_total_75perc,
    Age               = W1_mean_age,
    Ethnicity         = w1_ethnicity
    )

Summary Stats Table

datasummary_skim(df_names,
                 title = "Table 1. Sample Descriptive Statistics")

	Unique	Missing Pct.	Mean	SD	Min	Median	Max
Reappraisal	31	2	27.4	6.6	6.0	28.0	41.0
Suppression	24	2	13.7	5.2	5.0	13.0	28.0
Depression	49	5	12.8	10.3	0.0	9.5	50.0
Age	141	1	11.6	0.8	10.0	11.7	13.1
age	893	15	13.8	2.0	10.0	13.7	19.0
Ethnicity	N	%
a. Black/ African American	8	0.7
b. Hispanic/ Latino/ Chicano	48	4.2
c. Native American or Native Alaskan	8	0.7
d. White / Caucasian	720	63.4
e. Asian	16	1.4
g. Multi-racial	248	21.8
h. Other	16	1.4
j. Decline to respond	8	0.7
NA	64	5.6

Primary RQs

Is higher emotion reappraisal associated with lower depression symptoms in adolescent girls?

rq1 <- lm(df_tidy$ces_dc_total_75perc ~ df_tidy$erq_reappraisal_total)

Is higher emotion suppression associated with higher depression symptoms in adolescent girls?

rq2 <- lm(df_tidy$ces_dc_total_75perc ~ df_tidy$erq_suppression_total)

Exploratory RQs:

Is higher suppression use associated with higher depression symptoms when adjusting for reappraisal use?

ex1 <- lm(df_tidy$ces_dc_total_75perc ~ df_tidy$erq_suppression_total + df_tidy$erq_reappraisal_total)

Does age moderate the relation between reappraisal/suppression and depression symptoms in adolescent girls?

ex2 <- lm(df_tidy$ces_dc_total_75perc ~ df_tidy$erq_reappraisal_total * df_tidy$W1_mean_age)

ex3 <- lm(df_tidy$ces_dc_total_75perc ~ df_tidy$erq_suppression_total * df_tidy$W1_mean_age)

Primary RQ Model Summaries

Table 2. Model Summary for RQ1 and RQ2
	Reappraisal	Suppression
+ p < 0.1, * p < 0.05, p < 0.01, * p < 0.001
Intercept	22.805***	3.864***
	(1.306)	(0.800)
Reappraisal (ERQ)	-0.362***
	(0.046)
Suppression (ERQ)		0.645***
		(0.055)

Exploratory Results

Table 3. Model Summaries for Multiple Regression and Interaction Models
	Suppression + Reappraisal	Reappraisal × Age Interaction	Suppression × Age Interaction
+ p < 0.1, * p < 0.05, p < 0.01, * p < 0.001
Intercept	13.733***	40.986*	2.701
	(1.484)	(18.413)	(11.043)
Reappraisal (ERQ)	-0.332***	-1.688**
	(0.043)	(0.644)
Suppression (ERQ)	0.592***		0.053
	(0.055)		(0.743)
Reappraisal x Age		0.116*
		(0.056)
Suppression x Age			0.049
			(0.064)

Plot 1

This plot shows the relation between reappraisal and depressive symptoms

Plot 2

This plot shows the relation between suppression and depressive symptoms

Reflections

Most of our time was spent cleaning/managing the data - the models were quick!
Pivoting can be challenging and renaming can help
Don’t forget to make sure your variables are numeric after pivoting

Future Directions

Investigate longitudinal association between emotion regulation and depressive symptoms among adolescent girls.
Examine potential moderators and mediators such as brain circuit, puberty stage, and socioeconomic factors.
Test the reproducibility of the results using other dataset with larger sample size.

References

Barendse, M. E. A., Vijayakumar, N., Byrne, M. L., Flannery, J. E., Cheng, T. W., Flournoy, J. C., Nelson, B. W., Cosme, D., Mobasser, A., Chavez, S. J., Hval, L., Brady, B., Nadel, H., Helzer, A., Shirtcliff, E. A., Allen, N. B., & Pfeifer, J. H. (2019). Study Protocol: Transitions in Adolescent Girls (TAG). Frontiers in Psychiatry, 10, 1018. https://doi.org/10.3389/fpsyt.2019.01018

Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. https://doi.org/10.1037/1089-2680.2.3.271

Gross, J. J. (2015). Emotion regulation: current status and future prospects. Psychological Inquiry, 26(1), 1–26. https://doi.org/10.1080/1047840X.2014.940781

Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2), 348–362. https://doi.org/10.1037/0022-3514.85.2.348

Guyer, A. E., Silk, J. S., & Nelson, E. E. (2016). The neurobiology of the emotional adolescent: From the inside out. Neuroscience & Biobehavioral Reviews, 70, 74–85. https://doi.org/10.1016/j.neubiorev.2016.07.037

Hankin, B. L., Mermelstein, R., & Roesch, L. (2007). Sex differences in adolescent depression: stress exposure and reactivity models. Child Development, 78(1), 279–295. https://doi.org/10.1111/j.1467-8624.2007.00997.x

Kessler, R. C., Angermeyer, M., Anthony, J. C., Graaf, R. D., Demyttenaere, K., Gasquet, I., Girolamo, G. D., Gluzman, S., Gureje, O., Haro, J. M., Kawakami, N., Karam, A., Levinson, D., Mora, M. E. M., Browne, M. a. O., Posada-Villa, J., Stein, D. J., Tsang, C. H. A., Aguilar-Gaxiola, S., … Üstün, T. B. (2007). Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative. World Psychiatry, 6(3), 168. https://pmc.ncbi.nlm.nih.gov/articles/PMC2174588/

Nelson, E. E., Jarcho, J. M., & Guyer, A. E. (2016). Social re-orientation and brain development: An expanded and updated view. Developmental Cognitive Neuroscience, 17, 118–127. https://doi.org/10.1016/j.dcn.2015.12.008

Silvers, J. A. (2022). Adolescence as a pivotal period for emotion regulation development. Current Opinion in Psychology, 44, 258–263. https://doi.org/10.1016/j.copsyc.2021.09.023

Weissman, M. M., Orvaschel, H., & Padian, N. (1980). Children’s symptom and social functioning self-report scales. Comparison of mothers’ and children’s reports. The Journal of Nervous and Mental Disease, 168(12), 736–740. https://doi.org/10.1097/00005053-198012000-00005

Final Project